Search

Copyright of Baghdad Science Journal is the property of Republic of Iraq Ministry of Higher Education & Scientific Research (MOHESR) and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Summary: New cognitive radio (CR) systems require high throughput and bandwidth. Hence, CR users need to detect wide frequency bands of the radio spectrum to exploit unused frequency channels. This paper proposes a new wideband spectrum sensing (WBSS) detection approach based on machine learning (ML) for scanning subchannels. The originality of the proposed approach is to detect spectrum opportunities using a narrowband spectrum sensing (NBSS) method‐based support vector machine (SVM) classification and two features: energy and goodness of fit (GoF). The simulation results show that the proposed WBSS approach‐based ML presents a higher probability of detection than the WBSS approach‐based conventional detectors, even at low signal‐to‐noise ratio (SNR). Finally, the software defined radio (SDR) implementation validates the proposed WBSS approach for real detection scenarios. [ABSTRACT FROM AUTHOR]

Accurately counting the number of sorghum seedlings from images captured by unmanned aerial vehicles (UAV) is useful for identifying sorghum varieties with high seedling emergence rates in breeding programs. The traditional method is manual counting, which is time-consuming and laborious. Recently, UAV have been widely used for crop growth monitoring because of their low cost, and their ability to collect high-resolution images and other data non-destructively. However, estimating the number of sorghum seedlings is challenging because of the complexity of field environments. The aim of this study was to test three models for counting sorghum seedlings rapidly and automatically from red-green-blue (RGB) images captured at different flight altitudes by a UAV. The three models were a machine learning approach (Support Vector Machines, SVM) and two deep learning approaches (YOLOv5 and YOLOv8). The robustness of the models was verified using RGB images collected at different heights. The R2 values of the model outputs for images captured at heights of 15 m, 30 m, and 45 m were, respectively, (SVM: 0.67, 0.57, 0.51), (YOLOv5: 0.76, 0.57, 0.56), and (YOLOv8: 0.93, 0.90, 0.71). Therefore, the YOLOv8 model was most accurate in estimating the number of sorghum seedlings. The results indicate that UAV images combined with an appropriate model can be effective for large-scale counting of sorghum seedlings. This method will be a useful tool for sorghum phenotyping. [ABSTRACT FROM AUTHOR]

Copyright of Przegląd Elektrotechniczny is the property of Przeglad Elektrotechniczny and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Learning disability is a condition usual amongst most populace due to poor phonological capability in humans making them impaired. One such neurological disorder is developmental dyslexia, a lack of reading and writing skills leading to difficulty in school education. The essential causes of developmental dyslexia are the consumption of more drug treatments during pregnancy, the over-the-counter purchase of medicines for minor ailments without the recommendation of physicians, and uncared-for head accidents during early life. The occurrence of this trouble is acute in India. Attempts were made by many to detect dyslexic children to reduce the intensity of this hassle. In this proposed effort, machine learning is used to locate significant styles characterizing people using EEG samples. A dataset is used for examination of developmental dyslexia, and classification is done using K nearest neighbor (KNN), decision tree, linear discriminant analysis (LDA), and support vector machine (SVM) to evaluate the performance. This piece of research work is done on MATLAB to provide results on simulation with classification accuracy of 90.76% for SVM, sensitivity of 89% for SVM, and LDA with 91.89% specificity for SVM providing optimum yield. [ABSTRACT FROM AUTHOR]

A monthly runoff prediction model(STL-VMD-SVM) based on a secondary decomposition using loess (STL) and variational mode decomposition (VMD) combined with a support vector machine(SVM)was proposed to address the nonlinear and non-stationary characteristics of runoff sequences. This model utilized STL to decompose the original runoff sequence into seasonal, trend, and residual terms of different frequencies and decomposed the residual term into IMFs through VMD. An SVM model was established to predict seasonal, trend, and IMFs. The sum of the predicted values of all IMFs was the predicted value of the residual term, and the product of seasonal, trend, and residual terms was the final predicted value of the original runoff series. Based on the monthly runoff time series of Heishiguan Station and Gaocun Station on the mainstream of the Yellow River in the Yiluo River Basin, an example application and universality evaluation were conducted, and compared with the BP neural network model and the long shortterm memory neural network model(LSTM). The results showed that for the runoff prediction of Heishiguan Station in the Yiluo River Basin, the NSE, MAPE, RMSE, and R in the validation period of the proposed model were 0. 977, 13. 705%, 0. 327 and 0. 991, respectively, and their prediction accuracy was better than that of the single model and the primary decomposition model. The secondary decomposition of STL-VMD could effectively improve the prediction accuracy of the model. The NSE, MAPE, RMSE, and R during the validation period in the runoff prediction at Gaocun Station on the mainstream of the Yellow River were 0. 979, 8. 509%, 3. 263, and 0. 989, respectively, which also achieved good prediction results. [ABSTRACT FROM AUTHOR]

Interactions between individuals and digital material have completely changed with the advent of the Metaverse. Due to this, there is an immediate need to construct cutting‐edge technology that can recognize the emotions of users and continuously provide material that is relevant to their psychological states, improving their overall experience. An inventive method that combines natural language processing adaptive content generation algorithms and neuro‐fuzzy‐based support vector machines natural language processing (SVM‐NLP) is proposed by researchers to meet this demand. With this merging, the Metaverse will be able to offer highly tailored and engaging experiences. Initially, a neuro‐fuzzy algorithm was developed to identify people's emotional moods from their physiological reactions and other biometric information. Fuzzy Logic and Support Vector Machine work together to manage the inherent ambiguity and unpredictability, which results in a more exact and accurate categorization of emotions. A key component of the ACGA is NLP technology, which uses real‐time emotional data to dynamically modify and personalize characters, stories, and interactive features in the Metaverse. The novelty of the proposed approach lies in the innovative integration of neuro‐fuzzy‐based SVM‐NLP algorithms to accurately recognize and adapt to users' emotional states, enhancing the Metaverse experience across various applications. The proposed method is implemented using Python software. This adaptive approach significantly enhances users' immersion, emotional involvement, and overall satisfaction within the augmented reality environment by tailoring information to their responses. The findings show that the SVM‐NLP emotion identification algorithm based on neuro‐fuzzy, has a high degree of accuracy in recognizing emotional states, which holds promise for creating a Metaverse that is more emotionally compelling and immersive. Stronger human–computer interactions and a wider range of applications, including virtual therapy, educational resources, entertainment, and social media networking, might be made possible by integrating SVM‐NLP. These sophisticated systems are around 92% accurate in interpreting the emotions. [ABSTRACT FROM AUTHOR]

The internet and technological advancements have facilitated faster communication and information sharing. However, cybercrime, including malware, phishing, and ransomware, remains a severe problem despite technical progress. Detecting the intrusion via Intrusion Detection System IDS in network communication and wireless networks WSN is a big challenge that grown with the rapid development of the technologies. The detection accuracy of the IDS mainly depends on the relevant features of the incoming data from the internet. Selecting the most relevant features within the optimal attributes is one of the primary stage of the machine learning and pattern recognition modules. Finding the feature subset from the present or existing features that will improve the algorithms' learning performance in terms of accuracy and learning time is the main goal of feature selection. Therefore, this paper proposes an accurate approach for intrusion detection in the network and WSN using machine learning methods include Chaotic Maps, Nomadic People Optimizer (NPO), and SVM. The proposed approach has five main stages which are: data collection, pre-processing, feature selection, classification, and evaluation. An improved version of NPO based on chaotic map called CNPO is proposed. The proposed CNPO uses chaotic maps to initialize the population and solution distribution. Besides, a proposed fitness function for CNPO based on SVM is proposed. The CNPO is employed for feature selection task by selecting only the most relevant features from the input dataset. The proposed approach evaluated using two datasets and achieve accuracy 99.96% and 99.98 for NSL-KDD, and WSN-DS respectively. [ABSTRACT FROM AUTHOR]

This study addresses the critical need for effective groundwater (GW) management in Muzaffarabad, Pakistan, amidst challenges posed by rapid urbanization and population growth. By integrating Support Vector Machine (SVM) and Weight of Evidence (WOE) techniques, this study aimed to delineate GW potential zones and assess water quality. This study fills the gap in applying advanced machine learning and geostatistical methods for accurate GW potential mapping. Eight thematic layers based on topography, hydrology, geology, and ecology were utilized to compute the GW potential model. Additionally, water quality analysis was performed on collected samples. The findings indicate that flat and gently sloping terrains, areas with an elevation range of 611 –687 m, and concave slope geometries are associated with higher GW potential. Additionally, proximity to drainage and high-density lineament zones contribute to increased GW potential. The results showed that 31.1% of the area had excellent GW potential according to the WOE model, whereas the SVM model indicated that only 20.3% fell in the excellent potential zone. Results showed that both models performed well in the delineating GW potential zones. Nevertheless, the application of the SVM method is highly recommended which will be benefited in GW resources management related to urban planning. The study also evaluates the spatial distribution of GW quality, with a focus on physical and chemical parameters, including electrical conductivity, pH, turbidity, total dissolved solids, calcium, magnesium, chloride, nitrate, and sulphate. Bacterial contamination assessment reveals that 76% of spring water samples (30 out of 39 samples) are contaminated with E.coli, raising public health concerns. Based on the chemical analysis of GW samples the study identified exceedances of WHO guidelines for calcium in two samples, magnesium in seven samples, sulphate in ten samples, and nitrate levels were below the WHO guideline across all samples. These results highlight localized chemical contamination issues that require targeted remediation efforts to safeguard water quality for public health. [ABSTRACT FROM AUTHOR]

Recognizing speech emotions is indeed a crucial aspect of human–computer interaction. However, developing a model that can accurately process multiple languages is one of the challenging tasks. The feature selection process plays a vital role in multilingual speech emotion recognition because it helps to reduce irrelevant features from each language, ultimately enhancing the performance of the model. This research aims to address this task in a more precise way. It achieves this by employing Grid Search based Principal Component Analysis and an ensemble voting classifier for multilingual speech emotion recognition. Here we mention three essential steps of recognizing emotion from a multilingual dataset. The first step involves feature extraction from speech signals, such as MFCC, root-mean-square, ZCR, flux, roll-off, Centroid, bandwidth, chroma, and fundamental frequency. The second step entails the selection of an essential feature subset by removing redundant and unnecessary features using Principal Component Analysis. We also utilize the Grid Search technique to determine the feature subset that would yield the highest accuracy. The third step encompasses SVM and Random Forest, that are widely recognized classifiers. Additionally, we propose an ensemble voting classifier. Our study compares the performance of these classifiers on three distinct corpora—RAVDESS, EMOVO, and SUBESCO with and without the feature selection strategy. The accuracy for RAVDESS EMOVO and SUBESCO dataset 74.30%, 79.66%, 87.64%, respectively. After comparing our proposed approach with other approaches mentioned in the literature survey, it became evident that our approach outperforms the rest. [ABSTRACT FROM AUTHOR]

A variety of reasons have made it more difficult for educators and tutors to anticipate students' performance. Numerous researchers have used various predictive models to identify students who may be at-risk of dropping out early. Additionally, these methods were used to forecast final semester grades based on various datasets. However, these prediction models still fall short of meeting educational management requirements. In this paper, we propose the deep learning (DL) based model named students academic performance prediction network (SAPPNet) to predict the students' grades. We consider the questionnaire-based Jordan University dataset which contains demographic information, usage of digital tools before and after COVID-19, sleep times before and after COVID-19, social interaction, psychological state, and academic performance. SAPPNet consists of spatial convolution modules which are designed to extract spatial dependencies includes categorical and numerical attributes that represent static features (gender, level/year, age, digital tools used before and after COVID-19, psychological condition using prolonged e-learning tools) and temporal module for temporal dependencies involves sequences that capture changes before and after COVID-19. Additionally, we also try to implement classical machine learning (ML) models including support vector machine, k nearest neighbor, decision tree, and random forest, and DL models named artificial neural network, convolutional neural network, long short-term memory, and students learning prediction network. Simulation results show that SAPPNet achieved the best performance compared to state-of-the-art methods, with an accuracy, precision, recall, and an F1-score of 93 % . The proposed model with spatial and temporal modules improves the prediction performance, and it implies new aspect of the educational dataset. [ABSTRACT FROM AUTHOR]

Background/Objectives: Photoplethysmography (PPG) signals, which measure blood volume changes through light absorption, are increasingly used for non-invasive cardiovascular disease (CVD) detection. Analyzing PPG signals can help identify irregular heart patterns and other indicators of CVD. Methods: This research involves a total of 41 subjects sourced from the CapnoBase database, consisting of 21 normal subjects and 20 CVD cases. In the initial stage, heuristic optimization algorithms, such as ABC-PSO, the Cuckoo Search algorithm (CSA), and the Dragonfly algorithm (DFA), were applied to reduce the dimension of the PPG data. Next, these Dimensionally Reduced (DR) PPG data are then fed into various classifiers such as Linear Regression (LR), Linear Regression with Bayesian Linear Discriminant Classifier (LR-BLDC), K-Nearest Neighbors (KNN), PCA-Firefly, Linear Discriminant Analysis (LDA), Kernel LDA (KLDA), Probabilistic LDA (ProbLDA), SVM-Linear, SVM-Polynomial, and SVM-RBF, to identify CVD. Classifier performance is evaluated using Accuracy, Kappa, MCC, F1 Score, Good Detection Rate (GDR), Error rate, and Jaccard Index (JI). Results: The SVM-RBF classifier for ABC PSO dimensionality reduced values outperforms other classifiers, achieving the highest accuracy of 95.12% along with the minimum error rate of 4.88%. In addition to that, it provides an MCC and kappa value of 0.90, a GDR and F1 score of 95%, and a Jaccard Index of 90.48%. Conclusions: This study demonstrated that heuristic-based optimization and machine learning classification of PPG signals are highly effective for the non-invasive detection of cardiovascular disease. [ABSTRACT FROM AUTHOR]

Accurately counting the number of sorghum seedlings from images captured by unmanned aerial vehicles (UAV) is useful for identifying sorghum varieties with high seedling emergence rates in breeding programs. The traditional method is manual counting, which is time-consuming and laborious. Recently, UAV have been widely used for crop growth monitoring because of their low cost, and their ability to collect high-resolution images and other data non-destructively. However, estimating the number of sorghum seedlings is challenging because of the complexity of field environments. The aim of this study was to test three models for counting sorghum seedlings rapidly and automatically from red-green-blue (RGB) images captured at different flight altitudes by a UAV. The three models were a machine learning approach (Support Vector Machines, SVM) and two deep learning approaches (YOLOv5 and YOLOv8). The robustness of the models was verified using RGB images collected at different heights. The R² values of the model outputs for images captured at heights of 15 m, 30 m, and 45 m were, respectively, (SVM: 0.67, 0.57, 0.51), (YOLOv5: 0.76, 0.57, 0.56), and (YOLOv8: 0.93, 0.90, 0.71). Therefore, the YOLOv8 model was most accurate in estimating the number of sorghum seedlings. The results indicate that UAV images combined with an appropriate model can be effective for large-scale counting of sorghum seedlings. This method will be a useful tool for sorghum phenotyping. [ABSTRACT FROM AUTHOR]

Stress is a crucial factor that causes various health-related issues. Recent research has focused on stress prediction using Electroencephalography (EEG) signal processing. Early detection of stress among the students helps to avoid suicidal thoughts and illness, also proper counseling is given to improve the learning ability of the students. To improve the performance metrics of the classifier model, EEG features such as relative sub-band energies and EEG band ratios were considered. In this study, two levels of classification such as stress and non-stress states were carried out using machine learning techniques. An experimental work with EEG signal acquired from 25 subjects under two conditions as relax mode (non-stress) and during a mental task (stress) using an 8-channel wireless Enobio device. EEG features extracted using discrete wavelet transform technique, relative sub-band energy such as alpha, theta, and beta energies, and the relative band ratios computed from sub-band energies for two states such as arousal index, heart rate, performance enhancement index, cognitive performance attentional resource index (CPARI), CNS arousal and desynchronization. EEG Features were selected by analyzing statistically significant (p < 0.05) for both states of data by using a non-parametric test as the Wilcoxon signed-rank test, and brain functional connectivity analysis was carried out for subband energies. Then, the extracted features were imported to various machine learning classifier algorithms such as decision tree, linear discriminant analysis, Naïve Bayes, Support Vector Machine (SVM), k-Nearest Neighbor (kNN), ensemble, and neural network. The classifier performance metrics such as classification accuracy, sensitivity, specificity, and precision were compared for the above classifiers. The experimental result shows that the cubic SVM classifier has achieved the highest accuracy of 95.83%, sensitivity of 96.70%, specificity of 93.10% and precision of 97.78% for detecting stress and non-stress states compared with other classifier algorithms. A classification model was exported for the cubic SVM classifier and tested with an online EEG dataset for 12 subjects with two states as relaxed and during a task. Finally, the result for the exported cubic SVM classifier model was achieved with a classification accuracy of 89.74%. [ABSTRACT FROM AUTHOR]

We consider the problem of anomaly detection in proportional data by investigating the Libby-Novick Beta-Liouville distribution, a novel distribution merging the salient characteristics of Liouville and Libby-Novick Beta distributions. Its main benefit, compared to the typical distributions dedicated to proportional data such as Dirichlet and Beta-Liouville, is its adaptability and explanatory power when dealing with this kind of data. Our goal is to exploit this appropriateness for modeling proportional data to achieve great performance in the anomaly detection task. First, we develop generative models, namely finite mixture models of Libby-Novick Beta-Liouville distributions. Then, we propose two discriminative techniques: Normality scores based on selecting the given distribution to approximate the softmax output vector of a deep classifier and an improved version of Support Vector Machine (SVM) by suggesting a feature mapping approach. We demonstrate the benefits of the presented approaches through a variety of experiments on both image and non-image datasets. The results demonstrate that the proposed anomaly detectors based on the Libby-Novick Beta-Liouville distribution outperform the classical distributions as well as the baseline techniques. [ABSTRACT FROM AUTHOR]

The study aimed to develop a portable electronic nose system for detecting adulteration with soybean protein isolate (SPI) in chicken meat. The system mainly consisted of three parts: the gas sensor array, the DSP28335 control board, and the upper computer. The DSP28335 control board, developed using C language, included analog to digital converter (ADC) module, digital output (DO) module, pulse width modulation (PWM) module, controller area network (CAN) module, power module, drive circuit, and so forth. The upper computer, developed using LabVIEW, facilitated user interaction with the user by primarily handling CAN configuration and monitoring, displaying and storing sensor data, temperature and flow data, and sending and monitoring electronic nose commands. The feasibility of the proposed electronic nose for characterizing adulterated chicken meat was tested on six classes of chicken meat that had been adulterated with varied quantities of SPI. The mass fractions of SPI were 0%, 5%, 10%, 15%, 20%, and 25%, respectively. On the basis of odor data from the electronic nose, K‐nearest neighbor (KNN), linear discriminant analysis (LDA), and support vector machine (SVM) were applied to qualitatively distinguish minced chicken meat with different adulteration ratios. The results showed that the SVM model had the best recognition effect. When the best parameters (c, g) were c = 16 and g = 1, the accuracy of SVM model was 97.22% and 93.75% in the training and testing sets, respectively. These results demonstrated that the portable electronic nose designed in this paper effectively identifies minced chicken meat under various adulteration conditions, enabling rapid and nondestructive detection of chicken meat adulteration. [ABSTRACT FROM AUTHOR]

Electroencephalography (EEG) helps to assess the electrical activities of the brain so that the neuronal activities of the brain are captured effectively. EEG is used to analyze many neurological disorders, as it serves as a low-cost equipment. To diagnose and treat every neurological disorder, lengthy EEG signals are needed, and different machine learning and deep learning techniques have been developed so that the EEG signals could be classified automatically. In this work, five ensemble models are proposed for EEG signal classification, and the main neurological disorder analyzed in this paper is epilepsy. The first proposed ensemble technique utilizes an equidistant assessment and ranking determination mode with the proposed Enhance the Sum of Connection and Distance (ESCD)-based feature selection technique for the classification of EEG signals; the second proposed ensemble technique utilizes the concept of Infinite Independent Component Analysis (I-ICA) and multiple classifiers with majority voting concept; the third proposed ensemble technique utilizes the concept of Genetic Algorithm (GA)-based feature selection technique and bagging Support Vector Machine (SVM)-based classification model. The fourth proposed ensemble technique utilizes the concept of Hilbert Huang Transform (HHT) and multiple classifiers with GA-based multiparameter optimization, and the fifth proposed ensemble technique utilizes the concept of Factor analysis with Ensemble layer K nearest neighbor (KNN) classifier. The best results are obtained when the Ensemble hybrid model using the equidistant assessment and ranking determination method with the proposed ESCD-based feature selection technique and Support Vector Machine (SVM) classifier is utilized, achieving a classification accuracy of 89.98%. [ABSTRACT FROM AUTHOR]

Accurately classifying road signs is crucial for autonomous driving due to the high stakes involved in ensuring safety and compliance. As Convolutional Neural Networks (CNNs) have largely replaced traditional Machine Learning models in this domain, the demand for substantial training data has increased. This study aims to compare the performance of classical Machine Learning (ML) models and Deep Learning (DL) models under varying amounts of training data, particularly focusing on altered signs to mimic real-world conditions. We evaluated three classical models: Support Vector Machine (SVM), Random Forest, and Linear Discriminant Analysis (LDA), and one Deep Learning model: Convolutional Neural Network (CNN). Using the German Traffic Sign Recognition Benchmark (GTSRB) dataset, which includes approximately 40,000 German traffic signs, we introduced digital alterations to simulate conditions such as environmental wear or vandalism. Additionally, the Histogram of Oriented Gradients (HOG) descriptor was used to assist classical models. Bayesian optimization and k-fold cross-validation were employed for model fine-tuning and performance assessment. Our findings reveal a threshold in training data beyond which accuracy plateaus. Classical models showed a linear performance decrease under increasing alteration, while CNNs, despite being more robust to alterations, did not significantly outperform classical models in overall accuracy. Ultimately, classical Machine Learning models demonstrated performance comparable to CNNs under certain conditions, suggesting that effective road sign classification can be achieved with less computationally intensive approaches. [ABSTRACT FROM AUTHOR]

The onset of the COVID-19 pandemic has compelled a swift transformation in higher-education methodologies, particularly in the domain of course modality. This study highlights the potential for artificial intelligence and machine learning to improve decision-making in advanced engineering education. We focus on the potential for large existing datasets to align institutional decisions with student and faculty preferences in the face of rapid changes in instructional approaches prompted by the COVID-19 pandemic. To ascertain the preferences of students and instructors regarding class modalities across various courses, we utilized the Cognitive Process-Embedded Systems and e-learning conceptual framework. This framework effectively delineates the task execution process within the scope of technology-enhanced learning environments for both students and instructors. This study was conducted in seven Iranian universities and their STEM departments, examining their preferences for different learning styles. After analyzing the variables by different feature selection methods, we used three ML methods—decision trees, support vector machines, and random forest—for comparative analysis. The results demonstrated the high performance of the RF model in predicting curriculum style preferences, making it a powerful decision-making tool in the evolving post-COVID-19 educational landscape. This study not only demonstrates the effectiveness of ML in predicting educational preferences but also contributes to understanding the role of self-regulated learning in educational policy and decision-making in higher education. [ABSTRACT FROM AUTHOR]

The integration of different Mobile Edge Computing (MEC) applications has significantly enhanced the realm of security and surveillance, with Human Activity Recognition (HAR) standing out as a crucial application. The diverse sensors found in smartphones have made it convenient for monitoring applications to gather and analyze data, rendering them valuable for HAR purposes. Moreover, MEC can be employed to automate surveillance, allowing intelligent monitoring of restricted areas to identify and respond to unwanted or suspicious activities. This research develops a system using motion sensors in smartphones to identify unusual human activities. People's smartphones were employed to monitor both suspicious and regular activities. Information was collected for various actions categorized as either suspicious or regular. When a person performs a certain action, the smartphone records a series of sensory data, analyse important patterns from the basic data, and then determines what the person is doing by combining information from different sensors. To prepare the data, information from different sensors was aligned to a shared timeline. In this study, we used a sliding window approach on synchronized data to feed sequences into LSTM and CNN models. These models, which include initial layers of LSTM and CNN, automatically find important patterns in the order of human activities. We combined SVM with the features extracted by the shallow Neural Network to make a mixed model that predicts suspicious activities. Lastly, we compared LSTM, CNN, and our new shallow mixed neural network using a new real‐time dataset. The mixed model of CNN and SVM achieved an accuracy of 94.43%. Additionally, the sliding window method's effectiveness was confirmed with a 4.28% improvement in accuracy. [ABSTRACT FROM AUTHOR]

This study explores the application of a hybrid algorithm based on Differential Evolution (DE) and Support Vector Machine (SVM) in the mineral processing of elutriation machine. To address the problems of low quality and efficiency in metal beneficiation during elutriation, the DE-SVM algorithm was proposed, and a corresponding beneficiation quality prediction model was constructed. Experimental results showed that the average prediction accuracy and precision of the DE-SVM algorithm were 93. 7 % and 95. 6 %, respectively. The predicted concentrate recovery rate and the absolute error of predicted concentrate grade using the model were 98. 4 % and 0. 309 %, respectively. Compared with other algorithms and models, the DE-SVM algorithm and its associated elutriation machine beneficiation quality prediction model demonstrated significant advantages, providing an effective method to improve the quality and efficiency of precious metal beneficiation. [ABSTRACT FROM AUTHOR]

Corn syrup is a cost-effective sweetener ingredient for the food industry. In producing syrup from corn, process control to enhance and/or maintain a constant dextrose equivalent value (DE) is a constant challenge, especially in semi-automated/batch production settings, which are common in small to medium-size factories. Existing work has focused on continuous process control to keep parameter values within a setpoint. The machine learning method applied is for time series data. This study focuses on building process control models to enable semi-automation in small to medium-size factories in which the data are not as time dependent. Correlation coefficients were used to identify key process parameters that contribute to feed pH value and DE. Artificial neural network (ANN), support vector machine (SVM), and linear regression (LR) models were built to predict feed pH and DE. The results suggest (1) model accuracy ranges from 91% to 96%; (2) the ANN models yielded about 1% to 3% higher accuracy than the SVM and LR models and the prediction accuracy is robust even with as few as six data sets; (3) both the SVM and ANN models have noise tolerant properties, but ANN has a higher noise tolerance than SVM; (4) SVM performance can be hindered when using high-dimensional data sets; (5) the LR model yields higher variation in accuracy prediction than ANN and SVM; (6) distribution fitting is a good approach for generating data; however, fidelity of fitting can greatly impact accuracy; and (7) multi-stage models yield higher accuracy than single-stage models, but there are pros and cons to each approach. [ABSTRACT FROM AUTHOR]

In the field of autonomous driving, it is important to protect vulnerable road users (VRUs) and ensure the safety of autonomous driving effectively by improving the detection accuracy of VRUs in the driver's field of vision. However, due to the strong temporal similarity between pedestrians and cyclists, the insensitivity of the traditional least squares method to their differences results in its suboptimal classification performance. In response to this issue, this paper proposes an algorithm for classifying pedestrian and cyclist targets based on the micro-Doppler effect. Firstly, distinct from conventional time-frequency fusion methods, a preprocessing module was developed to solely perform frequency-domain fitting on radar echo data of pedestrians and cyclists in forward motion, with the purpose of generating fitting coefficients for the classification task. Herein, wavelet threshold processing, short-time Fourier transform, and periodogram methods are employed to process radar echo data. Then, for the heightened sensitivity to inter-class differences, a fractional polynomial is introduced into the extraction of micro-Doppler characteristics of VRU targets to enhance extraction precision. Subsequently, the support vector machine technique is embedded for precise feature classification. Finally, subjective comparisons, objective explanations, and ablation experiments demonstrate the superior performance of our algorithm in the field of VRU target classification. [ABSTRACT FROM AUTHOR]

The global shift towards sustainable energy has positioned photovoltaic (PV) systems as a critical component in the renewable energy landscape. However, maintaining the efficiency and longevity of these systems requires effective fault detection and diagnosis mechanisms. Traditional methods, relying on manual inspections and standard electrical measurements, have proven inadequate, especially for large-scale solar installations. The emergence of machine learning (ML) and deep learning (DL) has sparked significant interest in developing computational strategies to enhance the identification and classification of PV system faults. Despite these advancements, challenges remain, particularly due to the limited availability of public datasets for PV fault detection and the complexity of existing artificial-intelligence (AI)-based methods. This study distinguishes itself by proposing a novel AI-based approach that optimizes fault detection and classification in PV systems, addressing existing gaps in AI-driven fault detection, especially in terms of thermal imaging and current–voltage (I-V) curve analysis. This comprehensive survey identifies emerging trends in AI-driven PV fault detection, highlights the most advanced methodologies, and proposes a novel AI-based approach to enhance fault detection and classification capabilities. The findings aim to advance the state of technology in this field, offering insights into more efficient and practical solutions for PV system fault management. [ABSTRACT FROM AUTHOR]

In the context of energy conversion from renewable sources to distribution grids (insulated or not), a converter is often required to transfer energy from a low voltage source towards three-phase grids. This paper presents the HW design, the simulation results, and the conversion performance of a CSI converter intended to interface low-voltage renewable sources to three-phase grids. The main focus of this paper is to obtain the best performance in terms of voltage increase towards the output stage while maximizing the conversion efficiency. In comparison with the currently used energy conversion systems for small photovoltaic systems, hereafter some solutions were adopted to level and maximize the energy flow from the source to the DC-link and improve the quality of current supplied in terms of harmonic distortion. The proposed system is composed of two conversion stages: the first, voltage-to-current, the second current-to-current via a three-phase CSI bridge modulated with the SVM technique. The stages are not completely decoupled from an electrical point of view; therefore, in order to mitigate the effects of these interactions, synchronization strategies have been adopted. [ABSTRACT FROM AUTHOR]

Gastrodia elata Blume from different regions varies in growth conditions, soil types, and climate, which directly affects the content and quality of its medicinal components. Accurately identifying the origin can effectively ensure the medicinal value of G. elata Bl., prevent the circulation of counterfeit products, and thus protect the interests and health of consumers. Attenuated total reflectance Fourier transform infrared (ATR‐FTIR) spectroscopy is a rapid and effective method for verifying the authenticity of traditional Chinese medicines. However, the presence of scattering effects in the spectra poses challenges in establishing reliable discrimination models. Therefore, employing appropriate scattering correction techniques is crucial for improving the quality of spectral data and the accuracy of discrimination models. This study uses two ensemble preprocessing approaches; the first type is series fusion of scatter correction technologies (SCSF), and another method is sequential preprocessing through orthogonalization (SPORT). Four discriminant models were established using a single scattering correction technique and two ensemble preprocessing approaches. The results show that the data‐driven version of the soft independent modeling of class analogy (DD‐SIMCA) model built based on multiplicative scatter correction (MSC) preprocessing has a sensitivity of 0.98 and a specificity of 0.91, able to effectively distinguish whether a sample of G. elata Bl. originates from Zhaotong. In addition, three discriminant models including support vector machine (SVM), partial least squares discriminant analysis (PLS‐DA), and three gradient boosting machine (GBM) algorithms built using the ensemble preprocessing approach have good classification and generalization capabilities. Among them, the SCSF‐PLS‐DA model has the best performance with 99.68% and 98.08% accuracy for the training and test sets, respectively, and F1 of 0.97; the SPORT‐SVM model achieved the second‐best classification ability. The results show that the ensemble preprocessing approach used can improve the success rate of G. elata Bl. geographical origin classification. There are differences in the chemical composition of Gastrodia elata Blume from different regions. This study used single scanning correction technology and two ensemble preprocessing approaches to process ATR‐FTIR spectroscopy and established discrimination model for identifying the origin of G. elata Blume. The results indicate that the discriminative model established using two ensemble preprocessing approaches has the best discriminative performance. [ABSTRACT FROM AUTHOR]

This work proposes an electronic equipment which determines the marbling grade in beef rib eye according to the American grading scale using digital image processing and machine learning, achieving an 88.89 % coincidence level with grading done by beef specialists. Existing solutions which use image processing usually require calibration methods due to working in non-controlled environments. Furthermore, they only acquire the fat distribution from the longissimus dorsi muscle with an approximate accuracy of 80 %, without referring the distribution to any quality standard. In this work, meat samples are placed in a food grade stainless-steel enclosure with a touch screen and a digital RGB camera. The device acquires an image of the rib eye, which is then analyzed using techniques such as adaptive histogram analysis based on the HSV color model, histogram peaks detection for grayscale thresholding and a linear Support Vector Machine (SVM). The SVM determines the marbling grade based on the American Standard and shows it via a graphical user interface. The classifier was compared with a k-Nearest Neighbors (kNN) and Random Forest (RF) models, to choose the one with the best performance for marbling grade prediction. The SVM and the kNN models obtained a better performance than RF in identifying the marbling level. The estimated American Standard grade was compared to gold standard reference tests performed by specialists from the National Agrarian University in Lima-Peru, where the SVM achieved the aforementioned 88.89 % coincidence level. [ABSTRACT FROM AUTHOR]

This study investigates the technical challenges of applying Support Vector Machines (SVM) for multi-class classification in network intrusion detection using the UWF-ZeekDataFall22 dataset, which is labeled based on the MITRE ATT&CK framework. A key challenge lies in handling imbalanced classes and complex attack patterns, which are inherent in intrusion detection data. This work highlights the difficulties in implementing SVMs for multi-class classification, particularly with One-vs.-One (OvO) and One-vs.-All (OvA) methods, including scalability issues due to the large volume of network traffic logs and the tendency of SVMs to be sensitive to noisy data and class imbalances. SMOTE was used to address class imbalances, while preprocessing techniques were applied to improve feature selection and reduce noise in the data. The unique structure of network traffic data, with overlapping patterns between attack vectors, posed significant challenges in achieving accurate classification. Our model reached an accuracy of over 90% with OvO and over 80% with OvA, demonstrating that despite these challenges, multi-class SVMs can be effectively applied to complex intrusion detection tasks when combined with appropriate balancing and preprocessing techniques. [ABSTRACT FROM AUTHOR]

The penetration of renewable energy sources, electric vehicles (EVs) and load dynamics, and network complexities often lead to nuisance tripping in gridconnected microgrids. Traditional protection methods fail to discriminate fault and other dynamic volatilities in the system. The paper presents a novel twolevel adaptive relay algorithm to avoid nuisance tripping in a grid-connected microgrid under varying grid dynamics. The novelty of the adaptive relay algorithm is that nuisance tripping is eliminated by precisely determining normal system-level dynamics at the first level using a phase deviation reference block. The first level determines the necessity for activating the second level, which consists of a detection scheme combining a multiclass support vector machine (SVM) and discrete wavelet transform (DWT). The hybrid DWTSVM methodology ensures effective fault diagnosis, adapting to variations in energy sources, load fluctuations, and fault scenarios. Real-time hardware-in-the-loop (HIL) simulation validates the system's effectiveness in dynamic microgrid environments. Extensive experiments on scenarios, including faults, fluctuations in renewable energy generation, and intermittent simulations of EV charging and capacitor switching, were conducted to test the efficacy of the adaptive relay algorithm. Finally, experiments using OPAL-RT HIL realtime simulator and the Raspberry Pi microcontroller validated the adaptive relay algorithm in a grid-connected microgrid under varying grid dynamics. [ABSTRACT FROM AUTHOR]

Copyright of China Petroleum Machinery is the property of China Petroleum Machinery Editorial Department and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Recent research has suggested that category theory can provide useful insights into the field of machine learning (ML). One example is improving the connection between an ML problem and the design of a corresponding ML algorithm. A tool from category theory called a Kan extension is used to derive the design of an unsupervised anomaly detection algorithm for a commonly used benchmark, the Occupancy dataset. Achieving an accuracy of 93.5% and an ROCAUC of 0.98, the performance of this algorithm is compared to state-of-the-art anomaly detection algorithms tested on the Occupancy dataset. These initial results demonstrate that category theory can offer new perspectives with which to attack problems, particularly in making more direct connections between the solutions and the problem's structure. [ABSTRACT FROM AUTHOR]

This work combines the Support Vector Machine (SVM) model with Ensemble Empirical Mode Decomposition (EEMD) to present a novel method for drought prediction. The EEMD-SVM model is assessed for drought forecasting, comparing it to conventional Auto Regressive Integrated Moving Average (ARIMA) and SVM models. This study uses monthly precipitation data from Bamyan province in Central Afghanistan, spanning January 1970 to December 2019, including Standardized Precipitation Index (SPI) timelines: SPI 3, SPI 6, SPI 9, and SPI 12. To evaluate predictive effectiveness, statistical measures such as R-squared (R²), mean absolute error (MAE), and root-mean-square error (RMSE) are used. Each SPI series is decomposed into Intrinsic Mode Functions (IMFs) and a residual series using the EEMD approach. The next stage projects each IMF component and residual using the appropriate SVM model. The final step creates an ensemble forecast for the original SPI series by combining the anticipated results of the residual series with the modeled IMFs. Compared to traditional ARIMA and SVM models, results show that the EEMD-SVM technique greatly improves drought forecasting accuracy, especially for mid- and long-term SPI. For example, in the testing period, SPI 9 yields an RMSE of 0.1632, MAE of 0.1208, and R² of 0.9357, while for SPI 12, RMSE is 0.1078, MAE is 0.0745, and R² is 0.9141, indicating the best criteria with the lowest RMSE and MAE and highest R² compared to conventional ARIMA and SVM. This novel technology could enhance the capacity to forecast drought episodes, leading to more efficient water resource management and climate adaptation plans. [ABSTRACT FROM AUTHOR]

In a changing environment, the increasing inconsistency of runoff series complicates the development of runoff forecasting models. Mutation tests have frequently been employed to assess runoff inconsistency, yet their integration with runoff forecasting is rare. In this study, we proposed a combination of machine learning models based on mutation tests to predict monthly runoff at the Pingshi Station, located in the Lechang Gorge Reservoir of the Pearl River in Guangdong Province, China. Specifically, the mutation points of the monthly runoff were assessed using the Mann-Kendall test and the Moving T-test (MTT). The development of member models, including Artificial Neural Networks (ANN) and Support Vector Machines (SVM), utilized both the original runoff series and the sub runoff series identified after the first mutation point. Tele-connected factors and historical monthly runoffs served as candidate input variables. These variables were selected through linear and nonlinear filter methods and further refined by a greedy search based on 10-fold cross-validation. The improvement in overall forecasting performance by combining the ANN, SVM, and the Simple Average Method (SAM) was analyzed. The results showed that, for the validation dataset, the root mean square errors of the combined models with MTT decreased by about 14–19% compared to the member models with MTT, and 10–12% compared to the combined models without mutation tests. The corresponding Nash-Sutcliffe efficiency coefficients increased by approximately 35–45% and 12–14%, respectively. This study highlights the effectiveness of integrating mutation tests, machine learning, and combining models to enhance the forecasting performance of inconsistent monthly runoff. [ABSTRACT FROM AUTHOR]